Slashdot Mirror

Firefox has come a long way performance wise. Flexbox layout has been improved dramatically, and lot's of other performance improvements were implemented.

But there are still some major performance issues, and somehow I get the feeling that the Mozilla developers get a bit carried away with implementing new and exciting stuff instead of making the existing functionality really good.

If you are running Windows you may want to run the following test with hardware performance enabled and disabled, and compare the displayed frame rates.

Performance test: https://codepen.io/anon/pen/wY...

On my machine hardware acceleration reduces the frame rate at least by a factor 4(!). This is not what I understand by "acceleration". Please, Mozilla devs, this can't be what you had in mind when you introduced hardware acceleration.

I agree with the time frame being somewhere between ~1989 - ~2000.

1989 = NextStep -- Very good
1990 = Windows 3.1 -- Ugh, no!
1995 = Windows 95 -- Decent

I.e. HTML5/CSS demo of Win95 scrollbars:
https://codepen.io/louh/pen/oZ...

there is no specific rule that eliminates it from being a "programming language".

There is a in fact a difference. No one uses it as a programming languages, except for purposes of discussions such as this one.

It didn't take me long to find this:
https://medium.mybridge.co/26-...
and this:
https://codepen.io/collection/...

They are using the term programming language for simplicity sake, and are not technically incorrect.

OK. http://codepen.io/anon/pen/yOO...

You don't even need a text editor if you use a site such as CodePen.

Sure; we have artificial neural network algorithms. Check out this letter-recognition (backpropagation) network using 80 neurons that I wrote in JavaScript during a boring Christmas vacation with my parents. (And it sucks- not because it's JavaScript, but it makes embarrassing mistakes, which are the fault of the huge string literal of neuron weights at the end of the code).

Biologically, the process with a real neuronal cell body reaching a certain (unpredictable) voltage and firing is extremely complex. The firing mechanism is an analogue process, unstable and unreliable (which is how it works). It produces a digital signal which has an unpredictable time lag (the axon length and density of boundaries between glial cells affect this) before it reaches synapses (cesspools of quantum indeterminism) and tickles dendrites of other cells. The development and positioning of cell processes (axons, dendrites, synaptic junctions) is a necessary consequence of learning, but these are affected by gene expression and are extremely hard to predict.

Still, given this entire messy system, people's thoughts, free will, and reactions to stimuli are much more deterministic than they realize. But I suspect that if you wanted to make a robot that acts like a person does, you would at the very least need a prolific stream of very high quality random numbers. Maybe you can simulate the brain of Stephen Hawking with a PRNG; I haven't tried. (I sure as hell wouldn't use JavaScript!)